Of Isothermal Lines, and the Distribution of Heat over the Globe. By Alex. de Humboldt. Abridged from the third volume of the Memoirs of the Society of Arcueil. The unequal distribution of heat over the globe is one of those phenomena which has been long known as a general fact, but which cannot be exactly ascertained, with respect to its particular laws, until we have more correct data furnished by observation and experiment. To furnish these data is the immediate object of this paper; they are deduced from a great number of facts which have not been published; and if they are not sufficient to enable us to form a correct theory, they may, at least, lay a foundation for it, and will be useful in pointing out to travellers those objects to which they ought especially to direct their attention. The distribution of vegetables and of organized beings in general depends upon the circumstances that are connected with latitude, longitude, and elevation; and of these one of the most important is atmospherical temperature. The means which the author enjoyed of making observations, during his residence in South America, have enabled him to establish some very valuable data, which could not have been obtained in any other situation, more especially those which require to be made at great heights above the level of the ocean. The highest point in Europe where any observations have been made, is the Hospice de St. Gothard, at about 6,400 feet above the sea; but in South America the town of Quito is about 9,000 feet, Huancavelica about 11,670 feet, and the mine of Santa Barbara as much as 14,400 feet above the sea, or more than double that of St. Gothard. In order to compare the results which were obtained in the equinoxial regions with the mean heat of temperate climates, it was necessary to find out different situations, at intervals of 10° of latitude, but on different meridians, the mean temperature of which had been accurately ascertained. These will form so many fixed points, through which the isothermal lines, or the lines of equal temperature, may pass. In collecting facts for ascertaining these stations, comparatively few of the numerous thermometrical observations that have been published could be employed. Many of the observations contradict each other; in many cases we do not know under what circumstances they were made; and we are frequently obliged to reject such as, in other respects, appear correct, because we are not acquainted with the absolute height of the place at which they were made. This is remarkably the case with almost the whole of Asia; and it is not a little singular that, while there are more than 500 stations in the equinoxial district of America, many of them mere villages, or even hamlets, the altitude of which has been exactly determined, we are ignorant of the height of Bagdat, Aleppo, Ispahan, Delhi, and many other large and ancient cities in the Old World. In comparing the temperatures of different places, it is, however, quite necessary either that they should be upon the same level, or that a proper allowance should be made for any difference which there may be in this respect. In the old continent the only good observations which can be employed to form our calculations are limited by the parallels of 30° and 70° of latitude, and by the meridians of 30° east and 20° west longitude; the extreme points of which are, the island of Madeira, Cairo, and Cape North; it comprises about [Formel] the circumference of the globe from east to west. There are many circumstances connected with Europe, partly depending upon its natural form and situation, and partly upon the state of its civilization, which have given a peculiar character to its climate, different from that of other regions in the same latitude. But as this has been the abode of men of science, they have considered the laws which regulate the temperature of that part of the world as what are applicable to all the others. In this, however, they have fallen into some considerable errors, the causes and the amount of which we must endeavour to discover and appreciate. It is not with the temperature of the atmosphere, and with the magnetism of the globe, as with those phenomena which are determined by a single cause, and may be considered as distinct from all disturbing causes. From their nature, they both depend upon many local circumstances, such as the constitution of the soil, and the disposition of the radiating surface of the globe; and it frequently requires much judgment and discrimination to decide what circumstances are to be taken into account, and what are not connected with the inquiry. The object is to ascertain the quantity of heat which every part of the globe receives annually, and which is of actual use to agriculture and the well-being of the inhabitants; not what depends solely upon the action of the sun, its height above the horizon, or the extent of its semi-diurnal arcs. The temperature of a climate depends both upon the action of the sun, and of various extrinsic causes; among these may be enumerated the mixture of the temperatures of different latitudes, produced by winds, the vicinity of the sea, the nature of the soil, the presence and peculiar form of mountains, and the existence of large tracts of snow or masses of ice. In distinguishing between the solar and the real climate, we must remember that the local causes which modify the sun's action are themselves only secondary causes, effects which the motion of this luminary produces in the atmosphere. Many of the local or disturbing causes are necessarily connected with the nature of heat, and are felt over every part of the globe. The ocean tends to equalize the temperature of all the different regions by the mobility which necessarily belongs to it, and the currents of warm air which always flow from the equator to the poles, tend to diminish the rigour of the countries of the north. In estimating the action of the sun alone on the earth, we must have recourse to theory: this will not express the actual height of the thermometer in different situations, but it will show the relations between the mean temperature of different regions. By comparing the results of the calculation, not to the mean deduced from observations made in different longitudes, but to the mean temperature of a single point taken at the surface of the earth, we may proceed in our examination of what depends upon the sun, and upon all the other influences, solar or not solar, local, or such as extend to considerable distances. In this way we shall be able to form an interesting comparison between theory and experience. For the first philosophical ideas upon the subject of solar heat, we are indebted to Halley. Marian afterwards extended our knowledge of solar action; but he fell into some considerable errors, which were rectified by Lambert, who instituted some important calculations, which, however, do not always accord with actual experience, and which, indeed, depend almost entirely upon mathematical principles. After him, the subject was taken up by Moyer and by Kirwan; and they proceeded more upon observation, at least upon the method of endeavouring to approximate to a true system by collecting observations, and employing them to correct the theory. In speaking of the authors who have added to our knowledge on the temperature of the different parts of the earth, we must not omit to mention Cotte, who has collected a great number of documents; but they are not reduced into any kind of method, and are not all of equal authority. These considerations lead us to the conclusion that, in investigating the distribution of temperature, it is important to distinguish between the results which are deduced from observation, and those which are derived from theory. We must collect all the authentic facts that can be obtained; and, after arranging them into regular order, we must submit them to what may be termed empirical laws. After having correctly ascertained the mean temperature of certain places, we may trace on the globe isothermal lines, which thus exhibit to the eye the relation of these places to each other. In determining the mean temperature of a particular spot, the old plan was to take the maximum and minimum temperature of the year, and to consider the middle number between them as the mean temperature; but this plan is obviously incorrect. De la Hire seems to have been the first who attempted to pursue a different method, and one founded on more just principles: observing the uniformity of the temperature of the vaults that are attached to the observatory at Paris, he proposed as a general fact, that the temperature of vaults was the mean temperature of the climate. The other method, that of the maxima and minima, continued, however, to be generally adopted; and, by multiplying the number of the observations, it was rendered more correct, but still liable to inaccuracy. Some of the latest observers have noted the thermometer three times each day, and then taken the mean of these as denoting the mean temperature of the day; others have adopted the plan of observing the thermometer at two periods in the day, which are considered as indicating the maximum and minimumsun-rise, and two hours after noon; while others, again, have satisfied themselves with observing the temperature at one period only, which has been found, by previous experience, to denote the mean temperature. Mem. Acad. Sciences, 1719, p. 4. By comparing a great number of observations made between 46° and 48° N. latitude, we find that at the hour of sun-set the temperature is very nearly the mean of that at sun-rise and two hours after noon. When, besides noting the maximum and minimum, we take a middle observation, we shall fall into an error, if we simply divide the sum of the observations by three, without attending to the duration of the particular temperatures, and the place which the middle observation occupies between the extreme terms of the series. The middle observation should be at least four or five hours from either of the others; but, upon the whole, the two observations of the extreme temperatures will give us more correct results. Some very valuable observations have been made by noticing the thermometer from hour to hour, at different seasons of the year, and in different latitudes, until we are able to fix upon a number which may indicate the mean of the day. Serene and calm weather has been chosen on these occasions, and the thermometer has been carefully examined in this way, both in the observatory at Paris, and under the equator. These observations have tended to confirm the opinion that has been mentioned above, that the temperature of the earth corresponds with the mean temperature of the atmosphere, the disturbing causes nearly counteracting each other. In expressing the results of the observations on mean temperatures that have been made in various situations, it is convenient not to employ the numbers that are derived from any particular scale, but to consider the equator as the standard to which all the rest are to be referred, and to denote them all by numbers which have an arithmetical relation to it. Having now ascertained the method of taking mean temperatures, and of reducing them to a general expression, we may proceed to examine the form of some of the isothermal lines. It has been long known that the temperatures are not the same in the same parallels, especially those in Europe and in America; but, from the facts that will be stated, we shall find that this difference is not so great as has been imagined. By constructing a table, in which we compare the mean temperature and the latitude of different places in the continents of Europe and America, we learn the amount of this difference, and we deduce from this the number of degrees of latitude which we must go northward in Europe in order to arrive at the same annual mean. From this train of observations, we find that the isothermal line, or band, which is considered as 32°, passes between Ulea, in Lapland, latitude 66° 68', and Table Bay, in Labrador, latitude 54°. The isothermal line, or band, of 41°, passes near Stockholm, latitude 60°, and St. George's Bay, in Newfoundland, latitude 48°. The isothermal line, or band, of 50°, passes through Belgium, latitude 51°, and near Boston, latitude 42° 30'. The isothermal line, or band, of 59°, passes between Rome and Florence, latitude 43°, and near Raleigh, in South Carolina, latitude 36°. The direction of these lines of equal temperature gives the following differences between the west of Europe and the east of America: The thermometrical numbers are all reduced from the centigrade scale to that of Fahrenheit. In the original, the numbers on the centigrade scale are, 0°, 5°, 10°, 15°, respectively. Latitude. Mean of West of Europe. Mean of East of America. Differences. 30° .......... 70·1° .......... 66·8° .......... 3·3 40 .......... 63·1 .......... 54·5 .......... 8·6 50 .......... 50·8 .......... 37·9 .......... 12·9 60 .......... 40·0 .......... 24·0 .......... 16·0 In advancing from the equator towards the north pole, the mean temperatures become less than that of the equator in the following proportion: Lat. Temp. Temp. From 0°--20° in the old contin. 35·6° in the new 35·6° 20 --30 ............... 39·2 ........ 42·8 30 --40 ............... 39·2 ........ 44·6 40 --50 ............... 44·6 ........ 48·2 50 --60 ............... 41·8 ....... 45·1 0 --60 ............... 72·2 ........ 88·0 Having traced the isothermal belts from Europe to the provinces of the New World, the next object will be to observe them in North America itself. There are two chains of mountains in this continent--the Alleghanys, and the Rocky Mountains; the first running N.E. and S.W., the latter N.W. and S.E., making nearly equal angles with the meridian, and enclosing the vast plains of Louisiana, Tennesee, and the state of Ohio. This country possesses a milder climate than the parallel latitudes in the Atlantic States, the winters being less severe, and the summers less hot; so that the isothermal lines remain parallel, or nearly parallel, to the equator, from the coast of the Atlantic to the east of the Mississipi, and the Misoury. Beyond the Rocky Mountains the climate is still milder; in New California, and along the northern parts of the western side of the continent, the temperature appears to be very nearly the same with what it is in similar latitudes on the western side of Europe. The isothermal lines will, therefore, be bent upwards in this part. When we pass from the west of Europe eastward, the isothermal lines are again curved downwards; but the few accurate observations which we possess render it very difficult to fix the exact line; of the general fact, however, there is no doubt. We have hitherto found that, towards the north, the isothermal lines are neither parallel to the equator nor parallel to each other; and as the curve is the greatest in Asia and in America between 80° west and 100° east longitude, it might be supposed that the torrid zone of this part commences to the south of the tropic of Cancer, or that its heat is less intense. This, however, does not appear to be the case; as we approach to the line, below the parallel of 30°, the isothermal lines gradually become parallel to themselves and to the equator. For some time the old continent was thought to be warmer between the tropics than the new; but more correct observations have shown that this is not the case. The mean temperature of the equator may be fixed at 81·2.° The distribution of the temperature through the different parts of the year differs in the same isothermal line; this is the case with respect both to the old and new continents; in the former a few months are warmer than in the latter; as, for example, the heats in Madras are greater than those in Cumana. In the temperate zone it has been long known that the cold of the winter augments in a more rapid progression than the heat of the summer decreases; it is also known that the climate of islands, and the sea-coast, is milder than the interior of continents; it is, therefore, an important object to compare the mean temperatures of the three winter and the three summer months, at different latitudes, and to observe how the curves of the isothermal lines modify these relations. By comparing together a tract on the west with one on the east side of the Atlantic, extending across the different isothermal lines, we find that the difference between the two seasons increases more in the transatlantic than in the cisatlantic district. But in both the districts the division of the temperature between the winter and summer months is such, that upon the line of 32° the difference is nearly double what it is upon the line of 68°. In tracing the same isothermal line from west to east, in order to observe the difference between the winters and the summers, we find that the difference is less near the convex summits of the lines than near the concave summits. The same causes which tend to raise up the lines towards the pole also tend to equalize the seasons. Europe may be regarded altogether as the western part of a great continent, and subject to all those influences which make the western sides of all continents warmer than the eastern. The same difference that we observe in the two sides of the Atlantic exists on the two sides of the Pacific; in the north of China the extremes of the seasons are much more felt than in the same latitudes in New California, and at the mouth of the Columbia. On the eastern side of North America we have the same extremes as in China; New York has the summer of Rome and the winter of Copenhagen; Quebec has the summer of Paris and the winter of Petersburgh. And, in the same way in Pekin, which has the mean temperature of Britain, the heats of summer are greater than those at Cairo, and the cold of winter as severe as that at Upsal. This analogy between the eastern coasts of Asia and of America sufficiently proves that the inequalities of the seasons depend upon the prolongation and enlargement of the continents towards the pole, and upon the frequency of the N. W. winds, and not upon the proximity of any elevated tracts of country. TABLE I. Places. March. April. May. June. Differences of temp. of the four months. M. temp. of year. First Group. Concave summits in America. Natchez, lat. 31° 28' 58·0° 66·2° 72·6° 79·2° 8·2° 6·4° 6·6° 64·8° Williamsburg 37 18 46·4 61·2 66·6 77·8 14·8 5·4 11·2 58·1 Cincinnati ... 39 0 43·7 57·4 61·2 70·8 13·7 3·8 9·4 53·8 Philadelphia . 39 56 44·0 53·6 61·8 72·4 9·6 8·2 10·6 53·6 New York ... 40 40 38·6 49·1 65·8 82·2 10·5 16·7 16·4 53·8 Cambridge ... 42 25 34·6 45·5 56·8 70·2 10·9 11·3 13·4 50·4 Quebec ...... 46 47 23·0 39·6 54·6 63·8 16·6 15·0 9·2 41·8 Nain ........ 57 0 6·8 27·5 37·0 43·8 20·7 9·5 6·8 26·4 Second Group. Convex summits in Europe. (A) Climate of the continent. Rome ....... 41° 53' 50·4 55·4 67·0 72·4 5·0 11·6 5·4 60·4 Milan ....... 45 28 47·8 51·0 65·2 70·6 3·2 14·2 5·2 55·8 Geneva ...... 46 12 39·6 45·6 58·1 62·2 6·0 12·5 4·1 49·2 Buda ........ 47 29 38·3 49·1 64·8 68·4 10·3 15·7 3·6 60·0 Paris ........ 48 50 42·2 48·2 60·0 64·4 6·0 11·8 4·4 60·0 Gottingen .... 51 32 34·2 44·1 57·8 63·2 9·9 13·7 5·4 46·7 Upsal ....... 59 61 29·4 39·8 48·8 58·0 10·4 9·0 9·2 41·9 Petersburgh .. 59 56 27·5 36·9 52·2 59·4 9·4 15·3 7·2 38·8 Umea ....... 63 50 23·0 34·2 43·7 55·0 11·2 9·5 11·3 33·2 Ulea ........ 65 0 14·0 26·0 41·0 55·0 12·0 15·0 14·0 35·0 Enontekies .. 68 30 11·4 26·6 36·5 49·4 15·2 9·9 12·9 27·0 (B) Climate of the coast. Nantes ...... 47° 13' 50·0 53·6 60·0 65·6 3·6 13·6 5·6 54·6 London ...... 51 30 44·2 49·8 56·4 64·2 5·6 6·6 7·8 51·6 Dublin ...... 53 21 41·9 45·8 51·8 55·6 3·9 6·0 3·8 48·4 Edinburgh ... 57 57 41·4 47·3 50·6 57·2 5·9 3·3 6·6 47·8 Cape North .. 71 0 24·8 30·0 34·0 40·1 5·2 4·0 6·1 32·0 Third Group. Concave summit of Asia. Pekin ....... 39° 54' 41·4 57·0 70·4 84·2 15·6 23·4 13·8 54·8 TABLE II. Places. Cisatlantic band, longitude 29° E. 20° W. Places. Transatlantic band, longitude 67° E. 97° W. Lat. Mean temp. of the Lat. Mean temp. of the Year. Winter. Summer. Year. Winter. Summer. (Pondicherry) 11° 35' 85·4° 77·0° 90·8° Cumana ..... 10° 27' 81·6° 81·3° 83·3° Cairo ....... 30 2 72·6 57·6 84·6 Havannah ... 23 10 77·7 80·4 79·9 Funchal ..... 32 37 68·4 63·8 72·5 Natchez ..... 31 28 64·8 48·6 79·0 Rome ....... 41 55 60·1 45·8 75·2 Cincinnati ... 39 6 53·6 32·9 73·0 Bourdeaux ... 44 50 56·5 42·0 70·9 Philadelphia . 39 56 53·1 32·2 73·8 Paris ....... 48 50 51·8 38·3 66·2 New York .. 40 40 53·9 30·0 79·0 Copenhagen . 55 41 45·6 31·0 62·6 Cambridge ... 42 25 50·4 34·0 70·4 Stockholm ... 59 20 42·2 26·0 61·8 Quebec ...... 46 47 41·6 15·0 68·0 Drontheim ... 63 24 39·7 24·0 61·3 Nain ........ 57 10 26·4 - 0·4 48·4 Umea ....... 63 50 31·0 13·4 54·4 Fort-Churchill 59 2 25·5 6·8 52·0 The following table shows how the annual heat is divided between the two seasons of winter and summer in all the different parts of the temperate zone. The observations are traced along the isothermal lines from W. to E., and those are preferred which are situated nearest to the most curved parts of the lines; the longitudes are taken from the meridian of Paris: Mean temperature of Isothermal lines from 32° to 68°. Winter. Summ. Isothermal lines of 68°. Long. 84° 30' W.; lat. 29° 30'. (Florida). ........ 53·6 .. 80·6 Long. 19° 19' W.; lat. 32° 37' (Madeira).......... 63·6 .. 72·0 Long. 0° 40' E.; lat. 36° 48' (N. of Africa). ...... 59·0 .. 80·6 Isoth. line of 63·5°. Long. 92° W; lat. 32° 30' (Mississipi)............. 46·4 .. 77·0 Long. 11° 51' E.; lat. 40° 50' (Italy). ............ 50·0 .. 77·0 Isoth. line of 59°. Long. 86° 30' W.; lat. 35° 30' (Basin of the Ohio). 39·2 .. 78·4 Long. 1° 2' E.; lat. 43° 30' (South of France). .... 44·6 .. 75·2 Isoth. line of 54·5°. Long. 87° W.; lat. 38° 30' (Amer. W. of the Alleghanys)........................................ 34·7 .. 75·2 Long. 76° 30' W.; lat. 40° (Amer. E. of the Alleghanys)........................................ 32·7 .. 77·0 Long. 3° 52' W.; lat. 47° 10' (West of France)..... 41·0 .. 68·0 Long. 7° E.; lat. 45° 30' (Lombardy). ............ 34·7 .. 73·4 Long. 114° E.; lat. 40° (Eastern Asia)............ 26·6 .. 82·4 Isoth. line of 50°. Long. 86° 40' W.; lat. 41° 20' (Amer. W. of the Alleghanys)................................... 31·1 .. 71·6 Long. 73° 30' W.; lat. 40° (Amer. E. of the Alleghanys)........................................ 30·2 .. 73·4 Long. 9° W.; lat. 52° 30' (Ireland). .............. 39·2 .. 59·8 Long. 3° W.; lat. 53° 30' (England)..... ........ 37·4 .. 62·6 Long. 0°; lat. 51° (Belgium). .................... 36·5 .. 61·5 Long. 16° 40' E.; lat. 47° 30' (Hungary).......... 31·1 .. 69·8 Long. 114° E.; lat. 40° (Eastern Asia)............ 23·0 .. 78·8 Isoth. line of 45·5°. Long. 23° 20' W.; lat. 44° 42' (Amer. E. of the Alleghanys)........................................ 23·9 .. 71·6 Long. 4° 30' W.; lat. 57° (Scotland).............. 36·0 .. 56·4 Long. 10° 15' E.; lat. 55° 40' (Denmark).......... 31·3 .. 62·6 Long. 19° E.; lat. 53° 5' (Poland)................ 28·0 .. 66·2 Isoth. line of 41°. Long. 73° 30' W.; lat. 47° (Canada). ............ 14·0 .. 68·0 Long. 7° E.; lat. 62° 45' (Western Norway)....... 24·8 .. 62·6 Long. 15° E.; lat. 60° 30' (Sweden)............... 24·8 .. 60·8 Long. 22° E.; lat. 60° (Finland)................. 23·0 .. 63·5 Long. 34° E.; lat. 58° 30' (Centre of Russia)...... 13·0 .. 68·0 Isoth. line of 36·5°. Long. 74° W.; lat. 50° (Canada)................. 6·8 .. 60·8 Long. 15° 45' E.; lat. 62° 30' (W. coast of gulf of Bothnia)..................................... 17·6 .. 57·2 Long. 20° E.; lat. 62° 50' (E. coast of ditto)...... 16·7 .. 59·0 Isoth. line of 32°. Long. 60° W.; lat. 53° (Labrador). ............. 3·2 .. 51·8 Long. 17° 30' E.; lat. 65° (Sweden).............. 11·3 .. 53·6 Long. 23° E.; lat. 71° (Northern part of Norway). -- .. 45·7 These numbers, expressing the isothermal lines on the centigrade scale, are, 20°, 17 [Formel] °, 15°, 12 [Formel] °, 10°, 7 [Formel] °, 5°, 2 [Formel] °, 0°. We may perceive from this table that the inequality of the winters on the same isothermal line increases as the annual heat diminishes, from Algiers to Holland, and from Florida to Pennsylvania. If, instead of observing the most severe winter which is found in every climate, we trace the lines of similar winter temperatures, which we may style isocheimal lines; these, so far from coinciding with the isothermal lines, oscillate round them, and connect situations that are placed upon different isothermal lines. For example, in Belgium (geo. lat. 52°, isoth. lat. 51·8°), and even in Scotland (geo. lat. 57°, isoth lat, 45·5°), the winters are more mild than at Milan (geo. lat. 45° 28', isoth. lat. 57·7°). Ireland presents one of the most remarkable examples of the combination of very mild winters with cold summers; the mean temperature in Hungary for the month of August is 71·6°, while in Dublin it is only 60·8°. These, and many other instances which might be adduced, prove that the isocheimal lines vary much more from the terrestrial parallels than the isothermal lines; in the climates of Europe the latitude of two places, which have the same annual temperature, never differs more than 8° or 9°, while there are places that have the same winter temperature that differ in latitude 18° or 19°. The lines of equal summer heat, the isothermal curves, as we may style them, follow an exactly contrary direction to the isocheimal. We find the same summer heat at Moscow and at the mouth of the Loire, although the former is 11° further north than the latter; a circumstance which is attributed to the radiation of the earth in an extensive continent, without any considerable mountains. With respect to the relation which subsists between the temperature of winter and spring in different climates, it follows, from what has been stated above, that the increase of vernal temperature is considerable, and likewise much protracted, wherever the distribution of the annual temperature among the different seasons is very unequal, as in the north of Europe, and the more temperate part of the United States; that the vernal increase is great, but short, in the more temperate parts of Europe; that it is small, but protracted, in islands; and that in the different bands of climate enclosed between the same meridians, the vernal increase is smaller, and less protracted, in low than in high latitudes. Many very important conclusions are deduced from these facts respecting the effects of different climates on the cultivation of various kinds of plants, depending partly upon the absolute heat and cold of the summer and winter respectively, partly upon their relation to each other, and partly upon the transition from one season to the other. The southern hemisphere differs considerably from the northern; it is certainly colder; but the degree of difference between them has been very differently rated. The coldness of the southern hemisphere has generally been ascribed to the sun being a shorter space of time below than above the equator; but it probably depends more upon the greater proportion of ocean, which gives to the southern temperate zone a climate approaching to that of a collection of islands; there is, therefore, a less accumulation of heat during the summer, and a less radiation from the land, in proportion to its less extent; there is, consequently, a less current of warm air from the equator towards the south pole, which permits the ice to accumulate more round it. Near the equator, and indeed through the whole of the torrid zone, the temperature of the two hemispheres appears to be the same; but the difference begins to be felt in the Atlantic about the 22° of latitude; and there is a considerable difference between the mean temperature of Rio Janeiro and Havannah, although they are both equally distant from the equator, the former being 74·5°, the latter 76·4°. The southern climates generally differ from those of the north in respect to the distribution of the temperature through the different parts of the year. In the southern hemisphere, under the isothermal lines of 46° and 50°, we find summers which, in our hemisphere, belong to the lines of 35·5° and 41°. We are not accurately acquainted with the mean temperature of any place above 50° of south latitude; but there is every reason for supposing that it differs considerably from the same degree north. In estimating the temperature of the ocean there are four circumstances particularly to be attended to; 1. The temperature of the water at the surface, corresponding to the different latitudes, supposing it to be at rest, without either shallows or currents; 2. The decrease of heat in the strata of water which rest upon each other; 3. The effect of shallows, or banks, upon the temperature of the surface water; 4. The temperature of the currents which mix together the waters of different zones. The water of the ocean is said to be the warmest between 5° 45' N. and 6° 15' S.; it has been found, by different observers, to be from 82·5° to 84·5°; the temperature of the ocean in this part is from 4° to 6° higher than the temperature of the air which reposes upon it. As we advance towards the poles, the influence of the seasons upon the temperature of the surface of the sea becomes very sensible; but as a great mass of water follows the changes of the temperature of the air very slowly, the means of the months in the ocean and in the air do not correspond. To complete the subject of temperature, we have still to consider its variations in the different regions of the atmosphere, and in the interior of the earth; but our remarks have been already extended to so great a length, that we shall not, at present, enter upon these topics. Isothermal Bands, and Distribution of Heat over the Globe. The temperatures are expressed in degrees of Fahrenheit's thermometer; the longitudes are counted from east to west, from the first meridian of the observatory of Paris. The mean temperature of the seasons have been calculated so that the months of December, January, and February, form the mean temperature of the winter. The mark * is prefixed to those places the mean temperatures of which have been determined with the most precision, generally by a mean of 8000 observations. The isothermal curves having a concave summit in Europe, and two convex summits in Asia and Eastern America, the climate is denoted to which the individual places belong: Isothermal bands. Names of the places. Position in Mean temp. of the year. Distribution of Heat in the different Seasons. Maximum and Minimum. Latitude. Longitude. Height in feet. Mean temp. of winter. Mean temp. of spring. Mean temp. of summer. Mean temp. of autumn. Mean temp. of warmest month. Mean temp. of coldest month. Isothermal bands from 32° to 41°. Nain .......... 57° 8' 63° 40' W 0 26·8° - 0·4° 23·7° 48·4° 33·4° 51·8° - 11·2° 1 * Enontekies .... 68 30 18 27 E 1356 27·0 0·4 25·0 54·8 27·4 59·6 - 0·6 2 Hospice de St. Gothard... 46 30 6 3 E 6390 30·4 18·4 26·4 45·0 31·8 46·2 15·0 3 North Cape .... 71 0 23 30 E 0 32·0 23·8 29·4 43·2 32·2 50·2 22·1 4 * Ulea .......... 65 3 23 6 E 0 33·0 11·8 27·2 57·8 36·0 61·6 7·7 5 * Umea.......... 63 50 17 56 E 0 33·2 13·0 33·8 54·8 33·4 62·6 11·4 6 * Petersburg .... 59 56 27 59 E 0 38·8 17·0 38·2 62·0 38·6 65·6 8·6 7 Drontheim..... 63 24 8 2 E 0 40·0 23·8 35·2 61·4 40·1 65·0 19·8 8 Moscow. ...... 55 45 35 12 E 970 40·2 10·8 44·0 67·1 38·3 70·6 6·0 9 Abo. .......... 60 27 19 58 E 0 40·4 20·8 38·3 61·8 40·6 -- -- 10 Isothermal bands from 41° to 50°. * Upsal . ........ 59 51 15 18 E 0 42·0 25·0 40·0 60·2 42·8 62·4 22·4 11 * Stockholm ..... 59 20 15 43 E 0 42·2 25·6 38·3 61·8 43·2 64·0 22·8 12 Quebec ........ 46 47 73 30 W 0 41·8 14·2 35·9 68·0 46·0 73·4 13·8 13 Christiana..... 59 55 8 28 E 0 42·8 28·8 40·1 62·6 41·2 66·8 28·8 14 * Convent of Peyssenburg 47 47 8 14 E 3066 43·0 28·6 42·0 58·4 43·0 59·4 30·2 15 * Copenhagen.... 55 41 10 15 E 0 45·6 30·8 41·2 62·6 48·4 65·0 27·2 16 * Kendal ........ 54 17 5 6 W 0 46·2 36·8 45·2 56·8 46·2 58·1 34·8 17 Malouin Islands ...... 51 25 62 19 W 0 47·0 39·6 46·6 53·0 48·4 55·8 37·4 18 * Prague ........ 50 5 12 4 E 0 49·4 31·4 47·6 68·9 50·2 -- -- 19 Gottingen ...... 51 32 7 33 E 456 47·0 30·4 44·2 64·8 48·6 66·4 33·2 20 * Zurich ........ 47 22 6 12 E 1350 47·8 29·6 48·2 64·0 48·8 65·7 26·8 21 * Edinburgh. .... 55 57 5 30 W 0 47·8 38·6 46·4 58·2 48·4 59·4 38·3 22 Warsaw. ...... 52 14 18 42 E 0 48·6 27·8 47·4 69·0 49·4 70·4 27·2 23 * Coire ......... 46 50 7 10 E 1876 49·0 32·4 55·4 63·4 50·4 64·6 29·6 24 Dublin ........ 53 21 8 39 W 0 49·2 39·2 47·3 59·6 50·0 61·0 35·4 25 Berne. ........ 46 5 5 6 E 1650 49·3 32·0 49·0 66·6 49·8 67·2 30·6 26 * Geneva ........ 46 12 3 48 E 1080 49·3 34·9 47·6 65·0 50·0 66·6 34·2 27 * Manheim ...... 49 29 6 8 E 432 50·2 33·8 49·6 67·1 49·8 68·8 33·4 28 Vienna ........ 48 12 14 2 420 50·6 32·8 51·2 69·2 50·6 70·6 26·6 29 Isothermal band from 50° to 59°. * Clermont ...... 45 46 0 4 5 1260 50·0 34·7 50·6 64·4 51·2 66·2 28·0 30 * Buda .......... 47 29 16 41 494 51·0 31·0 51·0 63·2 52·4 71·6 27·6 31 Cambridge(U.S) 42 25 73 23 W 0 50·4 34·0 47·6 64·4 49·8 72·8 29·8 32 * Paris .......... 48 50 0 0 222 51·0 38·6 49·2 64·6 51·4 65·3 36·0 33 * London ........ 51 30 2 25 W 0 50·4 39·6 48·6 63·2 50·2 64·4 37·8 34 Dunkirk ...... 51 2 0 2 E 0 50·6 38·4 48·6 63·8 50·9 64·8 37·8 35 Amsterdam .... 52 22 2 30 E 0 51·6 36·8 51·6 65·8 51·6 67·0 35·4 36 Brussels. ...... 50 50 2 2 E 0 51·8 36·6 53·2 66·2 51·0 67·4 35·6 37 * Franeker ...... 52 36 4 2 E 0 51·8 36·6 51·0 67·2 54·4 69·0 32·9 38 Philadelphia. .. 39 56 77 36 W 0 53·4 32·2 51·4 74·0 56·6 77·0 32·7 39 New York. .... 40 40 76 18 W 0 53·8 29·8 51·2 79·2 54·6 80·6 25·4 40 * Cinninnati. .... 39 6 85 0 W 510 53·8 32·9 54·4 72·8 54·4 74·3 30·2 41 St. Malo. ...... 48 39 4 21 W 0 54·4 42·2 52·2 66·0 55·8 67·0 41·8 42 Nantes ........ 47 13 3 52 W 0 55·0 40·4 54·5 68·6 55·6 70·6 38·0 43 Pekin. ........ 39 54 114 7 E 0 55·2 26·8 56·3 82·6 54·2 84·4 39·4 44 * Milan. ........ 45 28 6 51 E 390 55·8 36·4 56·1 73·0 56·8 74·6 36·2 45 Bourdeaux .... 44 50 2 54 W 0 56·4 42·0 56·8 70·8 56·3 72·8 41·0 46 Isothermal band from 59° to 63°. Marseilles ..... 43 17 3 2 E 0 59·0 45·5 57·6 72·5 60·0 74·6 44·4 47 Montpelier .... 43 36 1 32 E 0 59·4 44·0 57·0 75·8 61·0 78·2 42·0 48 * Rome. ........ 41 53 10 7 E 0 60·4 45·8 57·8 75·2 62·8 77·0 42·2 49 Toulon........ 43 7 3 30 E 0 62·0 48·4 60·8 74·8 64·4 77·0 46·4 50 Nangasachi .... 32 45 127 35 E 0 60·8 39·4 57·6 83·0 64·2 86·9 37·4 51 * Natchez. ...... 31 28 93 50 W 180 64·8 48·6 65·4 79·2 65·8 79·7 47·0 52 Isothermal band from 68° to 77°. * Funchal. ...... 32 37 19 16 W 0 68·6 64·8 65·8 72·5 72·4 75·6 64·2 53 Algiers ....... 36 48 0 41 E 0 70·0 61·4 65·6 80·2 72·5 82·8 60·0 54 Isothermal bands above 77°. * Cairo. ........ 30 2 28 58 E 0 72·4 58·4 73·6 85·1 70·5 85·8 55·8 55 * Veracruz...... 19 11 98 21 W 0 77·8 72·0 77·9 81·5 78·6 81·5 71·0 56 * Havannah. .... 23 10 84 33 W 0 78·2 71·2 79·0 83·3 79·0 84·0 70·0 57 * Cumana. ...... 10 27 67 35 W 0 81·8 80·2 83·6 82·0 79·6 84·4 79·2 58 1 Coast of Labrador. Two years of observations. Floating ice towards the east. A transatlantic climate. Mean temperature of Oct. about 34·6°; Nov. 26·6°. 2 Centre of Lapland. A European climate. Fine vegetation. June, 49·4°; July, 59·6°; Aug. 56°; Sept. 41·8°; Oct. 27.5°; Nov. 12·4°. Inland situation. Specimen of a continental climate. 3 Eleven years of observations, calculated afresh in decads by Wahlenberg. Thermometer verified by Saussure. Mean temperature of seven months of the year below 32°. Winds blow from Italy in the winter. Minimum observed in the winter - 0·4°; in Aug. at noon, in the shade, maximum 54·5°; the nights in Aug. frequently from 33·8° to 29·3°; the mean temperature of Oct. 29·3° represents that of the whole year; at the Col de Geant, 10,598 feet high, the mean temperature of July is 36·5°. We find 32° to be the mean temperature in Europe, in 45° of latitude, at 5,400 feet high; at the parallel of the Canaries, at 12,300 feet; in the Andes, under the Equator, at 16,500 feet. 4 Buch, Voy. en Norw. ii. 416. Specimen of the climate of the islands and coasts in the north of Europe. April, 30°; May, 33·8°; Oct. 32°; Nov. 25·8°. At Alten, lat. 70°; mean temperature of July, 63·5°; a continental climate. 5 Finland, eastern coast. May, 40·8°; June, 55°; July, 61·6°; Aug. 56·6°; Sept. 46·6°; Oct. 38·6°; Nov. 24·6°. Julin and Buch. 6 Eastern coast of Western Bothnia. Dr. Noezen. March, 23·2°; April, 34°; Oct. 38·2°; Nov. 24·6°. 7 Euler. Mean temperature of the year, 38·2°. Inochodzow. Act. Petr. xii. 519--533. 8 Two years. Berlin, in the Mem. de l'Acad. de Drontheim, iv. 216. April, 34·4°; May, 50·8°; Oct. 39·2°; Nov. 27·8°. Climate of the west coast of Europe. 9 Four years. Journal de Phys. xxxix. 40. A continental climate. Winter colder, and summer warmer than at Petersburg. Eastern part of Europe; height as taken from Stritter. Chamounie, lat. 46° 1'; long. 3° 48' E.; height, 3,168 feet; mean temperature, 39·2°. 10 Twelve years. Kirwan. Cotte, mean of the year, 41·2°; of the summer, 67·4°, too high. West coast of Finland. 11 Observations from 1774 to 1804, made by Mallet, Prosperin, Holmquist, and Schleling, calculated by M. De Buch. Voy. de Norw. ii. 309. It is, perhaps, the place the mean temperature of which is the best determined. Winters more serene than at Stockholm; colder on account of the radiation of the ground and the air. 12 Thirty-nine years of observations, 15 of which are very good. Wargentin. Cotte, mean temperature of the year, 44·2°. Five months below 32°, as at Petersburg. 13 Four years. A transatlantic climate. 14 Buch, two years. Mean temperature of the winter scarcely 29·5°. West coast. 15 Alps of Bavaria. Six years' observations, calculated by M. Wahlenberg. Many fruit trees. Convent of Tegernsee, in Bavaria, height of 2,292 feet; mean temperature of 1785; 42·2°; Peyssenberg, 41°. 16 Bugge. Three months below 32°. Under the equator, mean temperature of 44·6°, at an elevation of 18,000 feet. 17 Dalton. West of England. Climate of islands; springs 47·2°. Keswick, lat. 54° 33', long. 5° 23' W.; mean temperature, 48°; springs, 48·6°. 18 Kirwan. Scarcely two years' observations. Southern latitude. 19 Strnadt. Fifteen years. Climate of the continent of Europe. 20 Maier. 21 Six years' observations of M. Escher, calculated by Wahlenberg. The town is situated in a hollow, to which the warm winds cannot penetrate, that render the winters more temperate in the other parts of Switzerland. 22 The calculation has been made from six years of excellent observations, by Professor Playfair; during this time the thermometer was never seen above 75·8°. Vegetation continues from March 20 to Oct. 20; mean temperature of these seven months is from 55·8° to 50·9°, according as the years are more or less fruitful; wheat does not ripen if the mean temperature descends to 47·6°. 23 Guittard. Only three years. Mean temperature a little too high. Eastern part of Europe. A continental climate. 24 Four years of observations, by M. de Salis Sewis, calculated by M. Wahlenberg. Mountains of the Grisons. 25 Kirwan. Irish Trans. viii. 203, and 269. Specimen of the climate of the islands. Coldest days, 23°; interior of the ground, 49·2°. Hamilton. 26 The climate of Berne is a continental climate, in comparison with that of Geneva; there is no lake near it. 27 Seven years of observations. Saussure. Mean temperature, 50·8°. Voy. § 1418. I find the mean temperature from 1796 --1815, 50°. Interior of the earth, 52°. Pictet, Bibliotheque Brit. 1817. iv. 109. 28 Six years. 29 Austria. Berlin, lat. 52° 31'; mean temperature, probably 46·4° to 47·3°; according to Beguelin, 48·8°; springs, 49·2°. Ratisbon, lat. 49°; height, 1,104 feet; mean temperature, 51·2°. Munich, lat. 48° 8'; height, 1,608 feet; mean temperature, 50·8°. 30 Ramond. Seven years of excellent observations. The mean of the months, at noon, well ascertained; winter, 40°; spring, 57°; summer, 70·8°; autumn, 58°. Mem. Inst. 1812. p. 49. Cotte, mean temperature, 51·2°. 31 Wahlenberg. Flor. Carp. p. 90. Continental climate. Height of the observatory, 474 feet. 32 Two years, near Boston, in New England. Transatlantic climate. The thermometer sometimes descends to 0°. 33 Eleven years (1803--1813) of observations made at the observatory. A greater number of years will, perhaps, give the mean temperature a little higher. Vaults, 53°. Kirwan finds for Paris, from seven years of observations of unequal value, 51·6°; he fixes upon 52·7°. Cotte, from 29 years of observations (Journ. de Phys. 1782, July), 53·2°. Cotte, for 33 years, (1763--1781, Mem. Instit. iv. 266), 52·4°. The extraordinary year of 1816 offers the mean temperature of 48·8°; winter, 37·2°; spring, 49°; summer, 59·6°; autumn, 50°: the preceding year, 1815, offers a mean temperature of 50·8°; winter, 37·2°; spring, 52·7°; summer, 62·8°; autumn, 50·8°. Arago. Mean temperature of Montmorency, for 33 years, 50·8°; height, 498 feet. Cotte. Strasburg, lat. 48° 34'; height, 480 feet; mean temperature, 49·2°. Herrenschneider. 34 Dr. Young. Mean temperature varies from 47·8° to 51·4°, (Lectures, ii. 453). Cavendish (Trans. 1788, p. 61), 48·8°, Roebuck, Hunter, and Kirwan, 51·6°. Horsley, 51·8°. According to Kirwan, the four seasons in London are, 39·6°, 50·9°, 64·8°, 52°; at Paris, 36·2°, 51°, 66°, 52·6°; from which results, London, 51·2°; Paris, 52·4°. Cotte (Journ. de Phys. xxxix. 36) thinks London is 51·2°, and Paris, 52·4°. The difference which we observe in cultivated plants depends less upon mean temperature than upon direct light, and the serenity of the atmosphere. 35 Seven years. Cotte. Lisle, 48·4°; Rouen, 51·4°; Cambray, 52°; Soissons, 53·4°; Rethel, 53·2°; Metz, 53°; Nancy, 52°; Etampes, 51°; L'Aigle, 49·8°; Brest, 54·2°; Mayenne, 52°. 36 Mohr, and Van Swinden. Five years. 37 Thirteen years. Temperature rather too high? 38 Eleven years. Van Swinden. From 1771--1783. Mean temperature, 51·2°. 39 Concave transatlantic summit. Seven years of observations give 54·8°; for the four seasons, 34°, 52°, 75·2°, 56·2°. Rush, 52·6 (Drake's View of Cincin. p. 116). Coxe, 54·2°. M. Legaux finds for 17 years, for Springmill on the Schuylkill, lat. 40° 50'; mean temperature, 53·4. Springs, near Philadelphia, 54·8°. Warden. 40 Two years only. Retif de la Serve. The thermometer sometimes descends to --4° in the parallel of Naples! Springs, 54·8°. Ipswich, lat. 42° 38'; mean temperature, 50°. Williamsburg, in Virginia, 58·1°. Cotte and Kirwan. Transatlantic climates. 41 Transatlantic climates west of the Alleghanys. Good observations, from 1806--1813. Col. Mansfield (Drake p. 93). Minimum of the winter, from 5° to 9·4°; Jan. 1797, as low as --16·6°, for 39° latitude. Maximum, 89·6° to 107·6° in the shade, without reflection; [Formel] of all the winds S.W.; springs near Cincinnati, 54·4°. Little snow falls; but it is abundant between lat. 40° and 42°. 42 Three years only. Bougourd. Dijon, height, 810 feet; lat. 47° 19'; mean temperature, 50·9°. Besancon, height, 804 feet; lat. 47° 14'; mean temperature, 51° 2'. 43 Six years. Duplessis, and Boudan. Temperature of the summer too high? Rochelle, 53°. Poitiers, 52·6°. 44 Amyot. Six years. Concave Asiatic summit. Three months below 32°, as at Copenhagen; the summer like that at Naples. 45 One of the best determined points. The years 1789--1812 are calculated in decads of days. Observations of the Astronomer Reggio, April, 55·8°; Oct. 58·1°. The two decads which approach the nearest to the mean temperature of the year, are, the first of April, 53·4°; and the last of Oct. 54·6°. The mean temperatures for January have varied in the last 10 years from 25° to 38·4°; those of July, from 71·4° to 78·4°; the mean of the years, from 54·5° to 57·2°. Reggio, taking only 24 maxima and minima in a year for 1763--1798; mean temperature, 55·4° (Ephem. Mil. 1779, p. 82). 46 Ten years. Guyot. Lyon, 528 feet, 55·8°. Mafra, near Lisbon, lat. 38° 52'; height, 600 feet; mean temperature, 54·3°, too small. Mem. de Lisbonne, ii. 105--158. 47 Seven years (1777--1782). St. Jaques de Sylvabella. The thermometer descends sometimes to 23°. Cotte (Traite de Met. ii. 420). 34 years (Raymond in Mem. de la Soc. de Med. 1777, p. 86) give 62°. Cotte (Journ. de Phys. xxxix. 21) fixes it at 58·6°. Kirwan, at 61·4°. The observations made at the Royal Observatory of Marseilles can alone decide. 48 Ten years. Nismes, 60·2°; Perpignan, 59·6; Tarascon, 60°; Arles, 59°; Rieux, 57·2°; Montauban, 55·6°; Tonains, 54·8°; Dax, 54·2°; Rodez, 57°; Aix, 56·6°. Under the equator, 57·8°, at 9,000 feet of elevation. 49 William Humboldt. Calandrelli, 60°. The thermometer sometimes descends to 36·5°, and rises to 99·5°. Naples, 67·1°; Toaldo, probably 63·5°; Florence, 61·6°; Tartini, too high; Lucca, 60·4°; Genoa, 60·2°; Bologna, 56·3°; Verona, 55·8°; Venice, 56·5°; Padua, 55·6°. Kirwan regards it as an established fact, that in Europe, the mean temperature, in latitude 40°, is 61·8°; in latitude 50°, 52·6°. 50 Only two years. Barberet, and d'Angos. Sheltered by mountains. Estimate a little too high. 51 Japan. A single year. Voy. de Thunberg, p. 121. Climate of islands. Under the equator, 64·4°, at a height of 6,000 feet. 52 West of the Alleghanys, in Louisiana. Four years. Dunbar. Transatlantic climate. 53 Madeira. Heberden. Climate of islands. St. Croix, of Teneriffe, 71·4°. The remainder of the island of Teneriffe, in the plains, 61·2°. Buch. 54 Old observations of Tartebout. They appear good. Bagdat, lat. 33° 19'; according to Beauchamps, 73·8°. The four seasons, 50·8°, 74·6°, 92·6°, 77°; but there was reflection from a house. The thermometer falls to 29·8°. Under the equator, at 3,000 feet high; mean temperature, 71·2°. 55 The calculations are made from the observations of Nouet (Decade, ii. 213). The following are the mean temperatures of the 12 months: 58·1°, 56·2°, 64·6°, 77·9°, 78·4°, 83·6°, 85·1°, 85·8°, 79·2°, 72·2°, 63°, 68·6°. (Neibuhr, 72·2.) Temperature of Joseph's Well, 72·5°. Catacombs of Thebes, 81·4°. Well of the great pyramid surrounded by sand, 88·2°. Jomard. Bassora, on the Persian Gulf; mean temperature, 77·9°; winter, 64°; summer, 90·8°; July, 93·2°. 56 Orta. Humboldt. Nouv. Esp. iv. 516. Jamaica, coast 80·6°. Blagden. 57 Ferrer, 1810--1812. Con. des Tems. 1817, p. 338. Wells of 10 feet deep; air, 76°; water, 74·4°; in 1812, maximum, Aug. 14, 86°; minimum, Feb. 20, 61·6°. Grottos, 81·5°. Humboldt, Observ. Astron. i. 134. 58 Humboldt. Pondicherry, 85·1°; Madras, 80·4°; Manilla, 78·2°; Isle de France, coast, 80·4°.